Objectives: Proposed is a randomized controlled study of four alternatives for the delivery of TB chemoprophylaxis with isoniazid (INH) to injection drug users (IDUs) recruited in street settings. The applicant seeks to evaluate the effectiveness of alternatives to traditional directly observed therapy and self-administered therapy using a randomized block design. Objectives are: (1) compare compliance with twice weekly medication schedules over a six month course across four models of TB chemoprophylaxis; (2) determine and compare cost- effectiveness of the four models; (3) identify demographic, behavioral, intervention related factors that are associated with compliance/non- compliance; (4) determine the prevalence of infection with Mycobacterium tuberculosis and active pulmonary TB in IDUs in nine neighborhoods in the San Francisco Bay Area; and (5) determine the prevalence of HIV infection in IDUs in East Oakland and in the Bayview-Hunter's Point/Visitation Valley section of San Francisco, and compare results with those of other neighborhoods examined in separately funded studies in the San Francisco Bay Area. Methods: Participants (N=404) will be drawn from a pool of verified IDUs participating in the P.I.'s ongoing HIV seroepidemiological surveys (N=2,000/yr). Participants with self-reported history of TB infection, who show positive Mantoux tests, or who show cutaneous anergy will be evaluated for active TB by collaborating TB clinics in the California counties of Alameda, Contra Costa, and San Francisco. Only latent cases for whom INH chemoprophylaxis is prescribed will be eligible for the compliance study component and will be randomly assigned to one of four modes of INH chemoprophylaxis. Modes of therapy delivery will be: (1) standard of care [or self-administered therapy] (SOC); (2) directly observed therapy in street and natural setting (DOT-S); (3) indirectly observed therapy in street and natural settings (I-DOTS); and (4) incentive-based directly observed therapy (IB-DOT). We will test the hypothesis: Compliance with SOC,I-DOTS, DOTS = IB-DOTS. Compliance will be measured as: (1) the proportion of therapy days completed/prescribed; and (2) the proportion of each group completing therapy. Statistical analyses include standard univariate comparisons (for descriptive purposes); the analysis of variance (for hypothesis testing); and logistic and multiple regression (to identify independent factors associated with compliance).